How Much Storage Do Security Cameras Need?

Quick answer

Security camera storage depends mainly on five inputs:

• Number of cameras
• Average recording bitrate
• Codec and video settings
• Continuous or motion-based recording schedule
• Required retention period

A practical estimate starts with this formula:

Storage (GB) = Camera count × Bitrate (Mbps) × 10.8 × Recording days × Recording duty cycle

The duty cycle is 1.0 for continuous recording. For motion recording, it represents the average percentage of time that cameras are actually recording—for example, 0.35 for 35% active recording time.

The result is an estimate of video data. Add capacity for filesystem overhead, camera metadata, audio, snapshots, changing bitrates, and free space. Then apply your chosen RAID layout to determine how many raw disk terabytes you need.

The inputs that determine surveillance storage

Camera count

More cameras increase both storage and write workload almost linearly.

If every camera uses the same bitrate and recording schedule:

Total storage = Storage per camera × Number of cameras

Cameras with different resolutions or settings should be calculated separately and then added together.

Average bitrate

Bitrate is usually the most useful input for capacity planning. It is commonly expressed in Mbps or Kbps.

Higher bitrate generally means:

• More image detail
• More storage consumed per day
• More sustained network traffic
• More write activity on the NAS or NVR

Use the camera’s average recording bitrate when available. A maximum or peak bitrate can produce a conservative estimate, but it may overstate normal storage use. Variable bitrate video can change significantly with motion, lighting, foliage, weather, and scene complexity.

Codec

H.264 and H.265 can produce different storage requirements at similar image quality, but codec efficiency depends on the camera, scene, frame rate, resolution, and encoder settings.

Do not estimate storage from codec name alone. Use the resulting bitrate from the camera or surveillance recorder whenever possible.

Codec choice can also affect playback compatibility and transcoding requirements. A storage estimate should use actual recorded bitrate, not a generic assumption about how much a codec “should” save.

Recording schedule

The two common recording modes are:

• Continuous recording: The camera records all day and night. Use a duty cycle of 1.0.
• Motion or event recording: The camera records only when motion or another event is detected. Use an estimated duty cycle based on actual activity.

Motion recording does not necessarily mean low storage use. Outdoor cameras may detect traffic, trees, rain, shadows, insects, or changing light. Event pre-recording and post-recording also add footage around each trigger.

If you do not have measured motion data, calculate both continuous and motion-based scenarios. Continuous recording provides a capacity ceiling; motion recording provides a scenario estimate rather than a guarantee.

The surveillance storage formula

For bitrate in Mbps:

Storage per camera per day (GB) = Bitrate (Mbps) × 10.8

This works because:

• 1 Mbps / 8 = 0.125 MB/s
• 0.125 MB/s × 86,400 seconds/day = 10,800 MB/day
• 10,800 MB/day = 10.8 GB/day using decimal units

For multiple cameras and a retention period:

Storage (GB) = Camera count × Bitrate (Mbps) × 10.8 × Days × Duty cycle

Or, in terabytes:

Storage (TB) = Camera count × Bitrate (Mbps) × 10.8 × Days × Duty cycle / 1,000

These are decimal storage calculations. Operating systems and storage applications may display capacity differently, often using binary units. Small differences between advertised capacity and displayed capacity are normal.

Formula for different camera bitrates

When cameras use different bitrates, calculate each group separately:

Total storage = (Cameras A × Bitrate A × 10.8 × Days × Duty cycle A) + (Cameras B × Bitrate B × 10.8 × Days × Duty cycle B)

This is more accurate than using a simple average when a few high-resolution cameras consume most of the storage.

Worked examples

Example 1: Eight cameras, continuous recording

Assume:

• 8 cameras
• 4 Mbps average bitrate per camera
• Continuous recording
• 30-day retention

8 × 4 × 10.8 × 30 × 1.0 = 10,368 GB

That is approximately:

10,368 GB / 1,000 = 10.37 TB

This is the estimated video data before adding free-space and operational allowances. A usable pool around this size would not be a sensible target because it leaves no room for variation or normal storage overhead.

If you plan a 20% capacity allowance:

10.37 TB × 1.20 = 12.44 TB

You would then choose a RAID layout and raw disk capacity that provides at least that much usable space.

Example 2: Eight cameras using motion recording

Use the same cameras, but assume they record for an average of 35% of the day:

8 × 4 × 10.8 × 30 × 0.35 = 3,628.8 GB

That is approximately 3.63 TB of estimated video data.

With a 20% planning allowance:

3.63 TB × 1.20 = 4.36 TB

This estimate is only as reliable as the 35% duty-cycle assumption. Review actual event volume after installation and adjust retention or capacity if necessary.

Example 3: Mixed camera types

Assume:

• 4 cameras at 2 Mbps, continuous, 30 days
• 2 cameras at 8 Mbps, continuous, 30 days
• 2 cameras at 4 Mbps, motion recording at a 40% duty cycle, 30 days

Calculate each group:

4 × 2 × 10.8 × 30 × 1.0 = 2,592 GB

2 × 8 × 10.8 × 30 × 1.0 = 5,184 GB

2 × 4 × 10.8 × 30 × 0.40 = 1,036.8 GB

Total estimated video:

2,592 + 5,184 + 1,036.8 = 8,812.8 GB

That is approximately 8.81 TB before a planning allowance and RAID overhead.

How resolution, frame rate, and scene affect storage

Resolution and frame rate influence bitrate, but they do not determine storage by themselves. Two cameras with the same resolution may use different bitrates because of encoder settings, image complexity, frame rate, and quality targets.

Storage use often rises when you increase:

• Resolution
• Frames per second
• Image quality or target bitrate
• Nighttime detail and noise
• Audio recording
• Keyframe frequency or other stream settings
• Number of recorded streams

Some systems record a high-quality main stream while using a lower-resolution substream for remote viewing or detection. Confirm which stream the NAS or NVR is actually storing.

Add a practical capacity allowance

Do not fill a NAS pool to its calculated video-data total. Keep capacity available for:

• Bitrate variation
• Motion bursts
• Pre-event and post-event recording
• Audio and snapshots
• Surveillance database files and metadata
• Filesystem and RAID overhead
• Temporary files and maintenance
• Future cameras or longer retention

A planning allowance of 15% to 25% is a reasonable starting point for many installations, but it is not a universal requirement. High-activity environments, uncertain motion recording, or future expansion justify more headroom.

Define the target as usable surveillance capacity, not the sum of advertised disk capacities.

Retention is a policy, not just a disk-size calculation

Decide what “retention” means before buying storage:

• Must every camera retain the same number of days?
• Are some cameras more important than others?
• Should continuous recording be used at entrances or cash-handling areas?
• Can lower-priority cameras use motion recording?
• Must recordings remain available during a network or NAS outage?
• Are legal, insurance, or organizational policies involved?

You can reduce capacity by assigning different recording policies to different cameras. For example, critical entrances may use continuous recording, while low-risk interior areas use motion recording. This saves space but creates a more complex system to configure and audit.

Write workload: can the NAS handle the cameras?

Storage capacity is only one part of surveillance sizing. The NAS must also accept the combined recording stream reliably.

Approximate aggregate write bandwidth as:

Aggregate bitrate (Mbps) = Sum of all camera recording bitrates

Convert it to megabytes per second:

Bandwidth (Mbps) / 8 = theoretical MB/s

For example, 16 cameras recording at 4 Mbps each produce:

16 × 4 = 64 Mbps

64 Mbps / 8 = 8 MB/s

This is the video payload only. Network protocols, file operations, metadata, playback, backups, snapshots, and other NAS services add overhead.

A surveillance workload is often a sustained write workload with many simultaneous streams. Check more than the headline sequential speed of a NAS:

• Number of simultaneous recording streams
• Supported surveillance applications or licenses
• Number and type of storage drives
• Network interface speed
• Concurrent playback and mobile viewing
• File-system and RAID overhead
• Other NAS services running at the same time
• What happens when the NAS is busy with backups or maintenance

A low aggregate bitrate does not guarantee a trouble-free installation if the NAS is also serving large files, running virtual machines, or handling many clients.

Network throughput and camera topology

The network must carry the camera streams from the cameras or PoE switch to the recording system.

For camera traffic:

Total camera bitrate = Number of streams × Average bitrate per stream

Add margin for protocol overhead and other network traffic. A dedicated or appropriately segmented surveillance network can make troubleshooting easier and reduce competition with normal user traffic.

Consider:

• PoE switch uplink capacity
• Uplink from the camera network to the NAS or NVR
• VLAN or network segmentation requirements
• Whether cameras send directly to the NAS
• Whether an NVR receives streams and writes them to storage
• Remote viewing and cloud-upload traffic
• Network behavior during camera reconnects or outages

Cameras may continue recording locally if they support onboard storage, but that behavior depends on the camera and recording system. Confirm the failover design rather than assuming the network will always be available.

RAID: useful for availability, not backup

RAID combines multiple drives to provide capacity, redundancy, performance, or a mixture of these goals. RAID does not protect against every cause of data loss.

Common layouts include:

• RAID 1: Mirrored data. Usable capacity is approximately half of the raw capacity of two equally sized drives.
• RAID 5: Single-drive parity. Usable capacity is approximately the total capacity of all drives minus one drive, subject to minimum-size and system overhead.
• RAID 6: Dual-drive parity. Usable capacity is approximately the total capacity of all drives minus two drives.
• RAID 10: Mirrored pairs striped together. Usable capacity is approximately half of raw capacity, with different fault-tolerance behavior from RAID 5 or RAID 6.
• ZFS redundancy: Capacity and protection depend on the vdev layout, such as mirrors, RAIDZ, or RAIDZ2. You must size the complete pool design rather than treating “ZFS” as a RAID level.

These are planning approximations. Mixed drive sizes, reserved space, RAID metadata, filesystem overhead, and the NAS implementation affect the final usable figure.

For surveillance, redundancy can keep recording available after a drive failure, but it also affects rebuild time, write behavior, and expansion options. Larger arrays may make drive replacement and recovery more consequential. Choose a layout based on the number of drives, required availability, rebuild risk, performance, and future expansion—not capacity alone.

RAID is not backup

RAID does not replace a backup. It will not reliably protect against:

• Accidental deletion
• Ransomware or compromised administrator credentials
• File corruption replicated across the array
• Fire, flood, theft, or power-related damage
• A failed NAS or controller
• Incorrect retention or deletion settings

A surveillance backup may include exported incident footage, configuration files, or selected recordings rather than every video stream. Define what must be recoverable and how quickly it must be restored. Keep at least one backup copy separate from the primary NAS, and test restoration.

NAS versus NVR for security camera recording

A NAS is a general-purpose storage server. Depending on its software and compatibility, it may:

• Store camera recordings
• Run surveillance software
• Provide user access and file sharing
• Run backups and other applications
• Support snapshots or replication
• Expand into broader home or business storage duties

An NVR is purpose-built around video surveillance. It commonly handles camera discovery, recording schedules, event indexing, playback, alerts, and camera management in one system.

The correct choice depends on the camera ecosystem and the features you need. A NAS may be attractive when you already need shared storage or want one platform for multiple workloads. An NVR may be simpler when surveillance is the primary requirement.

Do not assume that any NAS can record any camera. Check the NAS surveillance application, supported camera protocols, camera limits, recording licenses, codec support, playback features, and retention behavior before buying.

A hybrid design is also possible: an NVR manages cameras while a NAS stores exports or backups. This can separate surveillance management from general storage, but it adds configuration and network dependencies.

Do you need SSD cache for surveillance storage?

SSD cache is not automatically useful for camera recording. Surveillance writing is often sustained and sequential, while an SSD cache may be more valuable for metadata, random access, or other NAS applications.

Before adding cache, determine:

• Whether the NAS supports read cache, write cache, or both
• Whether a protected write cache is available
• Whether the surveillance application benefits from caching
• Whether the cache persists data safely during power loss
• Whether the NAS has enough memory and drive performance without it
• Whether cache activity competes with the actual recording workload

For many installations, correctly sized hard drives, adequate network capacity, reliable power protection, and a suitable RAID layout matter more than adding SSD cache. Treat cache as a workload-specific decision, not a substitute for insufficient storage capacity or an undersized network.

Power protection and operational reliability

A surveillance system is only useful if recordings remain intact and available during interruptions.

Consider:

• A UPS sized for the NAS, networking equipment, and any NVR
• Safe shutdown support
• Automatic restart after power is restored
• Drive health monitoring
• Alerts for failed disks, disconnected cameras, and low capacity
• Clock synchronization across cameras, NAS, and NVR
• Firmware and security updates
• Physical protection for the NAS and network equipment

A UPS does not provide indefinite runtime. Its purpose is to bridge short outages and allow an orderly shutdown when necessary.

Expansion planning

Plan for more than the current camera count if the system is likely to grow.

Future capacity may be needed for:

• Additional cameras
• Higher-resolution replacements
• Higher frame rates
• Longer retention
• Continuous recording replacing motion recording
• More exported incident footage
• Other NAS workloads

Expansion is not equally simple with every RAID or ZFS layout. Some systems support adding drives or expanding a pool, but the process can take time and may temporarily affect performance or protection. Check the NAS manufacturer’s documented expansion method before relying on it.

A practical expansion plan is to decide the target camera count, retention period, redundancy level, and maximum acceptable downtime before selecting the initial chassis and number of drive bays.

Surveillance storage sizing checklist

Use this checklist before purchasing a NAS, NVR, or drives:

• [ ] Count every camera that will record.
• [ ] Record the average bitrate for each camera or camera group.
• [ ] Confirm the codec and which stream is being stored.
• [ ] Note resolution, frame rate, audio, and image-quality settings.
• [ ] Decide which cameras record continuously.
• [ ] Estimate motion-recording duty cycles from real activity where possible.
• [ ] Set the required retention period in days.
• [ ] Calculate video data using Cameras × Mbps × 10.8 × Days × Duty cycle.
• [ ] Add a practical allowance for variation, metadata, and free space.
• [ ] Convert usable capacity into raw capacity for the selected RAID or ZFS layout.
• [ ] Check simultaneous write streams and aggregate network bandwidth.
• [ ] Confirm PoE switch and uplink capacity.
• [ ] Verify NAS surveillance software, camera compatibility, and licensing.
• [ ] Decide whether a NAS, NVR, or hybrid system better fits the workload.
• [ ] Plan backups for important recordings and configurations.
• [ ] Provide UPS protection and monitoring.
• [ ] Reserve capacity and drive bays for likely expansion.

If you are still selecting the storage platform, Browse NAS to compare available NAS options. For systems intended to handle surveillance alongside file sharing, backups, or other services, review NAS & storage servers.